Here you have the measurements and it actually seems to be a tweeter with a capable low end!
Tweeter group 8
Listening to the SB 29 & PE RS 28 down low....
OK...
Have used both of these tweeters... seeking an alternative to the SS D3806... a genuine tweeter that can play down low with authority... like 1,200 Hz flat... minus 6 DB give or take at 800 Hz.
Both of these are great tweeters... the measurements don't lie... but... run them at reasonable levels... even with resonance traps and they just do not hold up. Seems the lower limit... in terms of flat repsonse is limited to about 1350 Hz or so. They just run out of displacement. They just can not move enough air to play any lower in a real application. You hear displacement compression... no two ways about it... dynamics die quickly if you play them any lower. It is really obvious when you listen to them over time.
Just an emperical observation to be sure.... but that is my take. You just have to have the area and X-max to pull it off... be it a woofer or tweeter.
All of this has been realized as I have been testing many tweeters in search of a replacement for the SS D3806 in my Manzanita OB design.... have used wave guides... and steep crossovers and all the rest.... funny though.. it turns out that Scan Speak is no longer part of Tympany... they are now out on their own again... and guess what... they are making the D3806 again!!
So... my personal search now ends... but.... much has been learned and the SB and PE tweeters can be pushed to operate down to 1,350 Hz or so for other two way designs.
Mark... thanks for the testing and evaluation... oh so useful!!
Best
OK...
Have used both of these tweeters... seeking an alternative to the SS D3806... a genuine tweeter that can play down low with authority... like 1,200 Hz flat... minus 6 DB give or take at 800 Hz.
Both of these are great tweeters... the measurements don't lie... but... run them at reasonable levels... even with resonance traps and they just do not hold up. Seems the lower limit... in terms of flat repsonse is limited to about 1350 Hz or so. They just run out of displacement. They just can not move enough air to play any lower in a real application. You hear displacement compression... no two ways about it... dynamics die quickly if you play them any lower. It is really obvious when you listen to them over time.
Just an emperical observation to be sure.... but that is my take. You just have to have the area and X-max to pull it off... be it a woofer or tweeter.
All of this has been realized as I have been testing many tweeters in search of a replacement for the SS D3806 in my Manzanita OB design.... have used wave guides... and steep crossovers and all the rest.... funny though.. it turns out that Scan Speak is no longer part of Tympany... they are now out on their own again... and guess what... they are making the D3806 again!!
So... my personal search now ends... but.... much has been learned and the SB and PE tweeters can be pushed to operate down to 1,350 Hz or so for other two way designs.
Mark... thanks for the testing and evaluation... oh so useful!!
Best
D3806 at Madisound
Be sure to be sitting down before you look at the price - ouch!!
Fast and fun the project remains, but not inexpensive with that driver.
Be sure to be sitting down before you look at the price - ouch!!
Fast and fun the project remains, but not inexpensive with that driver.
The Manzanita SS D3806 solution
As many know this basic design has been around a while now... the original built on a whim of sorts to validate an idea I had about a viable stand mounted OB with a baffle size roughly equivalent to the larger Monkey Coffin designs of the 60's & 70's (large stand mounted 2 and 3 ways)... think AR3, Large Advent, Yamaha 1000 and JBL L-100. My goal was a simple system in terms of wood work, size, two way layout if possible, low cross over part count.... usable bass into the 30's, easy to drive, a reasonable efficiency (for a small OB), able to play at mid 90's db levels with 6 db reserve using off the shelf components. And of coarse keeping the raw materials price reasonable at less than $150 per system... plus some sweat equity.
The original design was a two way based around the venerable Scan Speak D3806. On the low end the new at the time Peerless 830669 looked to be perfect. But at that time the D3806 shot up from a wholesale value of $56 each to over $100 in the span of a few months and stock was limited. There were rumors that the purchase of Scan Speak by Tymphany could result in the 3806 being discontinued. So I switch to the also "venerable" Seas 27TDFC. It was lower in cost and with the correct crossover tricks could play safely down to about 1,300 Hz. I could make that work... but the D3806 could go down to about 850 Hz. Some crossover and performance changes and compromises were made to create the original, field-able Manzanita using the TDFC.
I have since move up to the D3806 and now that it is available again... Tymphany DID discontinue the D3806 but now that Scan Speak now owns Scan Speak again.... a crossover suitable for the D3806 is now of use. Keep in mind the D3806 is VERY expensive.... like a couple of hundred EACH by the time you get them home.
There is no one ideal solution... variables like the size of room, back wall distance, available power... near field vs. distant field listening distance preferences and so forth need to be considered in the final crossover configuration. Pretty common stuff. To my ears... the best over all utility set up... assuming a 48" distance from the rear wall and at least the same from any side wall(s)... with the speaker on a 24" pedestal.... I use the following crossover.
Woofer.... 16 Ga.. iron core inductor.... 20-22 MH in series.
D3608 Tweeter.....
Series Resonance trap across the tweeter terminals.... assuming a resonance of 370 Hz. which is typical: 60 to 70 mfd + 2.25 MH + 10 ohm.
Tweeter series padding resistor of 14 ohms ..... and series cap of 6.5 MFD and a shunt inductor of 8 MH to complete the 2nd order electrical... which with the tweeter's natural roll off becomes 3rd order Acoustic. You can vary the shunt inductor from 3 MH to 10 Mh to get the sound balance that is right for you. Keep in mind... the larger the room and farther from the wall the more balanced the sound and the deeper the bass. Like most OB's ... the Manzanita likes to breath.
As best I can measure you end up around 82 DB/watt... give or take.... and the system will need 100 clean watts per channel to come alive.... 200 is better. A lot of power to be sure... but with power you get some amazing dynamics ... bass dynamics for a small OB design. Some will like a hotter high end ... so the pad resistor can be lowered down to about 10 ohms. As you go down in pad you will need some downward adjustment for the series cap. Suspect many systems will end up at about 12 ohms.
That's it for now.... hope this answers most questions.
As many know this basic design has been around a while now... the original built on a whim of sorts to validate an idea I had about a viable stand mounted OB with a baffle size roughly equivalent to the larger Monkey Coffin designs of the 60's & 70's (large stand mounted 2 and 3 ways)... think AR3, Large Advent, Yamaha 1000 and JBL L-100. My goal was a simple system in terms of wood work, size, two way layout if possible, low cross over part count.... usable bass into the 30's, easy to drive, a reasonable efficiency (for a small OB), able to play at mid 90's db levels with 6 db reserve using off the shelf components. And of coarse keeping the raw materials price reasonable at less than $150 per system... plus some sweat equity.
The original design was a two way based around the venerable Scan Speak D3806. On the low end the new at the time Peerless 830669 looked to be perfect. But at that time the D3806 shot up from a wholesale value of $56 each to over $100 in the span of a few months and stock was limited. There were rumors that the purchase of Scan Speak by Tymphany could result in the 3806 being discontinued. So I switch to the also "venerable" Seas 27TDFC. It was lower in cost and with the correct crossover tricks could play safely down to about 1,300 Hz. I could make that work... but the D3806 could go down to about 850 Hz. Some crossover and performance changes and compromises were made to create the original, field-able Manzanita using the TDFC.
I have since move up to the D3806 and now that it is available again... Tymphany DID discontinue the D3806 but now that Scan Speak now owns Scan Speak again.... a crossover suitable for the D3806 is now of use. Keep in mind the D3806 is VERY expensive.... like a couple of hundred EACH by the time you get them home.
There is no one ideal solution... variables like the size of room, back wall distance, available power... near field vs. distant field listening distance preferences and so forth need to be considered in the final crossover configuration. Pretty common stuff. To my ears... the best over all utility set up... assuming a 48" distance from the rear wall and at least the same from any side wall(s)... with the speaker on a 24" pedestal.... I use the following crossover.
Woofer.... 16 Ga.. iron core inductor.... 20-22 MH in series.
D3608 Tweeter.....
Series Resonance trap across the tweeter terminals.... assuming a resonance of 370 Hz. which is typical: 60 to 70 mfd + 2.25 MH + 10 ohm.
Tweeter series padding resistor of 14 ohms ..... and series cap of 6.5 MFD and a shunt inductor of 8 MH to complete the 2nd order electrical... which with the tweeter's natural roll off becomes 3rd order Acoustic. You can vary the shunt inductor from 3 MH to 10 Mh to get the sound balance that is right for you. Keep in mind... the larger the room and farther from the wall the more balanced the sound and the deeper the bass. Like most OB's ... the Manzanita likes to breath.
As best I can measure you end up around 82 DB/watt... give or take.... and the system will need 100 clean watts per channel to come alive.... 200 is better. A lot of power to be sure... but with power you get some amazing dynamics ... bass dynamics for a small OB design. Some will like a hotter high end ... so the pad resistor can be lowered down to about 10 ohms. As you go down in pad you will need some downward adjustment for the series cap. Suspect many systems will end up at about 12 ohms.
That's it for now.... hope this answers most questions.
Yeah, that's spendy! May be why Madisound does so much overseas business.
I've seen NOS on eBay for $90. In fact John bought a few.
It is a very well built driver and just does so well with a 12-15" woofer. The harmonic structure seems to mate well with big woofers. Great piano sounds. I ran a pair with a cheap 12" in sealed boxes for years. Very nice. They ran 16 hours a day for years, often quite loud (huge space). They were universally liked.
I've seen NOS on eBay for $90. In fact John bought a few.
It is a very well built driver and just does so well with a 12-15" woofer. The harmonic structure seems to mate well with big woofers. Great piano sounds. I ran a pair with a cheap 12" in sealed boxes for years. Very nice. They ran 16 hours a day for years, often quite loud (huge space). They were universally liked.
Sorry, almost forgot – no, there is no formula as any working MinPhase contour for now has to be found by excessive BEM simulation by simple try and error.
Here is the file with the dipole horn co-ordinates for the B&G NEO3-W:
http://www.kinotechnik.edis.at/pages/diyaudio/DDCD/NEO3/DiHo_Neo3_381.csv
Playing around with the front / rear cut shape of top and bottom plate showed that we can get even better frequency response from that quasi MinPhase horn:
The contour in the form of a sliced fan developed by producing a lot of sawdust with the jig saw yesterday and hence is not "easy" to draw.
I'll hope to come back with a "cleaned up" version and a corresponding drawing (with comparable results from measurements)
As for the other questions – yes there is (and must be) a different vertical polar response of course – as compromises are largely a matter of taste I suggest to simply give it a try ...
I use the same approach with a big AMT for quite a while now and all I can say is it sounds excellent to me – if you stand or sit or sit even *out* of the stereo triangle (as I quite often do).
Sure – tonality and stereo image changes – but its still a pleasure to listen ...
Actually this cheap NEO3-W horn for the Manzanita is a spin off of my bigger system to simply allow the idea to spread...
Michael
Last edited:
Michael,
I don't want to unduly lengthen the discussion of your MinPhase at Lynn's thread. This one seems more appropriate to do so.
I have to correct myself and admit that above FR looks better than the Summa's indeed within the first 30 deg. Can you give any indication if FR develops in a similar way above that - may be up to 90 deg? I am looking for that good match with a OB dipole low-mid too.
And yes - your description of the new "sliced fan " shape may be detailed, but I am still lost like others. No chance to see any illustration of what you were doing?
Rudolf
I don't want to unduly lengthen the discussion of your MinPhase at Lynn's thread. This one seems more appropriate to do so.
I have to correct myself and admit that above FR looks better than the Summa's indeed within the first 30 deg. Can you give any indication if FR develops in a similar way above that - may be up to 90 deg? I am looking for that good match with a OB dipole low-mid too.
And yes - your description of the new "sliced fan " shape may be detailed, but I am still lost like others. No chance to see any illustration of what you were doing?
Rudolf
Some nice work going on here. Will be fun to see where Michael's experiences with the planar will lead. Soft termination of the waveguide/horn will certainly help, however it's done.
Back to the topic of baffles, both John and I have found that a soft covering like felt can make a big improvement to the sound of this speaker. Especially in the near-field. John is using a rough stone-like finish on his current pair and that also works well. The soft or rough baffle face seems to scatter and diffuse the reflections off the baffle face and that helps lower the noise floor of the speaker. Very worthwhile.
Back to the topic of baffles, both John and I have found that a soft covering like felt can make a big improvement to the sound of this speaker. Especially in the near-field. John is using a rough stone-like finish on his current pair and that also works well. The soft or rough baffle face seems to scatter and diffuse the reflections off the baffle face and that helps lower the noise floor of the speaker. Very worthwhile.
Well actually I would have hoped I'm in a *DIY* thread
but to be serious - thanks for the flowers – it is actually a lot of fun to play around with that cute little toy to bring the best out of such a limited (cylinder wave) design. Thankfully in audio its all about choosing the right compromises.
What could prove that any better than your beautiful Manzanita thread?
Thanks
I'm happy you corrected that –
I pretty exactly know how Earls OS performs from his own measurements my own simulations and the lengthy discussions spread over several threads.
It was late last night when I developed the "sliced fan" so I stopped measuring at 30deg – but will come back to you with up to 60deg as soon as the "sliced fan" is readily cleaned up. This may take a while as I will not have that much time for serious work the days to come.
Michael
Last edited: